Mirror of espurna firmware for wireless switches and more
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

354 lines
11 KiB

Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
Rework settings (#2282) * wip based on early draft. todo benchmarking * fixup eraser, assume keys are unique * fix cursor copy, test removal at random * small benchmark via permutations. todo lambdas and novirtual * fix empty condition / reset * overwrite optimizations, fix move offsets overflows * ...erase using 0xff instead of 0 * test the theory with code, different length kv were bugged * try to check for out-of-bounds writes / reads * style * trying to fix mover again * clarify length, defend against reading len on edge * fix uncommited rewind change * prove space assumptions * more concise traces, fix move condition (agrh!!!) * slightly more internal knowledge (estimates API?) * make sure cursor is only valid within the range * ensure 0 does not blow things * go back up * cursor comments * comments * rewrite writes through cursor * in del too * estimate kv storage requirements, return available size * move raw erase / move into a method, allow ::set to avoid scanning storage twice * refactor naming, use in code * amend storage slicing test * fix crash handler offsets, cleanup configuration * start -> begin * eeprom readiness * dependencies * unused * SPI_FLASH constants for older Core * vtables -> templates * less include dependencies * gcov help, move estimate outside of the class * writer position can never match, use begin + offset * tweak save_crash to trigger only once in a serious crash * doh, header function should be inline * foreach api, tweak structs for public api * use test helper class * when not using foreach, move cursor reset closer to the loop using read_kv * coverage comments, fix typo in tests decltype * ensure set() does not break with offset * make codacy happy again
4 years ago
  1. /*
  2. Part of the DEBUG MODULE
  3. Copyright (C) 2016-2019 by Xose Pérez <xose dot perez at gmail dot com>
  4. Copyright (C) 2019-2020 by Maxim Prokhorov <prokhorov dot max at outlook dot com>
  5. */
  6. // -----------------------------------------------------------------------------
  7. // Save crash info
  8. // Original code by @krzychb
  9. // https://github.com/krzychb/EspSaveCrash
  10. // -----------------------------------------------------------------------------
  11. #include "espurna.h"
  12. #if DEBUG_SUPPORT
  13. #include "crash.h"
  14. #include "system.h"
  15. #include "rtcmem.h"
  16. #include "storage_eeprom.h"
  17. #include <cstdio>
  18. #include <cstdarg>
  19. /**
  20. * Structure of the single crash data set
  21. *
  22. * 1. Crash time
  23. * 2. Restart reason
  24. * 3. Exception cause
  25. * 4. epc1
  26. * 5. epc2
  27. * 6. epc3
  28. * 7. excvaddr
  29. * 8. depc
  30. * 9. adress of stack start
  31. * 10. adress of stack end
  32. * 11. stack trace size
  33. * 12. stack trace bytes
  34. * ...
  35. */
  36. #define SAVE_CRASH_CRASH_TIME 0x00 // 4 bytes
  37. #define SAVE_CRASH_RESTART_REASON 0x04 // 1 byte
  38. #define SAVE_CRASH_EXCEPTION_CAUSE 0x05 // 1 byte
  39. #define SAVE_CRASH_EPC1 0x06 // 4 bytes
  40. #define SAVE_CRASH_EPC2 0x0A // 4 bytes
  41. #define SAVE_CRASH_EPC3 0x0E // 4 bytes
  42. #define SAVE_CRASH_EXCVADDR 0x12 // 4 bytes
  43. #define SAVE_CRASH_DEPC 0x16 // 4 bytes
  44. #define SAVE_CRASH_STACK_START 0x1A // 4 bytes
  45. #define SAVE_CRASH_STACK_END 0x1E // 4 bytes
  46. #define SAVE_CRASH_STACK_SIZE 0x22 // 2 bytes
  47. #define SAVE_CRASH_STACK_TRACE 0x24 // variable, 4 bytes per value
  48. static constexpr int EepromCrashBegin = EepromReservedSize;
  49. static constexpr int EepromCrashEnd = 256;
  50. static constexpr size_t CrashReservedSize = EepromCrashEnd - EepromCrashBegin;
  51. static constexpr size_t CrashTraceReservedSize = CrashReservedSize - SAVE_CRASH_STACK_TRACE;
  52. static constexpr uint32_t EmptyTimestamp { 0xffffffff };
  53. namespace debug {
  54. namespace {
  55. namespace crash {
  56. namespace internal {
  57. bool enabled = true;
  58. } // namespace internal
  59. namespace build {
  60. constexpr bool enabled() {
  61. return 1 == SAVE_CRASH_ENABLED;
  62. }
  63. } // namespace build
  64. namespace settings {
  65. bool enabled() {
  66. return getSetting("sysCrashSave", build::enabled());
  67. }
  68. } // namespace settings
  69. bool enabled() {
  70. return internal::enabled;
  71. }
  72. void enableFromSettings() {
  73. internal::enabled = settings::enabled();
  74. }
  75. size_t reserved() {
  76. if (enabled()) {
  77. return CrashReservedSize;
  78. }
  79. return 0;
  80. }
  81. // Simply reset the timestamp to stop dump() from printing the output more than once per crash.
  82. void clear() {
  83. eepromPut(EepromCrashBegin + SAVE_CRASH_CRASH_TIME, EmptyTimestamp);
  84. eepromCommit();
  85. }
  86. // Print out crash information that has been previusly saved in EEPROM
  87. // Optionally, check whether the timestamp is erased / EEPROM contains no data.
  88. void dump(Print& print, bool check) {
  89. char buffer[256] = {0};
  90. uint32_t crash_time;
  91. eepromGet(EepromCrashBegin + SAVE_CRASH_CRASH_TIME, crash_time);
  92. bool crash_time_erased = ((crash_time == 0) || (crash_time == EmptyTimestamp));
  93. if (check && crash_time_erased) {
  94. return;
  95. }
  96. uint8_t reason = eepromRead(EepromCrashBegin + SAVE_CRASH_RESTART_REASON);
  97. if (!crash_time_erased) {
  98. snprintf_P(buffer, sizeof(buffer), PSTR("\nlatest crash was at %lu ms after boot\n"), crash_time);
  99. print.print(buffer);
  100. }
  101. snprintf_P(buffer, sizeof(buffer), PSTR("Reason of restart: %u\n"), reason);
  102. print.print(buffer);
  103. if (reason == REASON_EXCEPTION_RST) {
  104. snprintf_P(buffer, sizeof(buffer), PSTR("\nException (%u):\n"),
  105. eepromRead(EepromCrashBegin + SAVE_CRASH_EXCEPTION_CAUSE));
  106. print.print(buffer);
  107. uint32_t epc1, epc2, epc3, excvaddr, depc;
  108. eepromGet(EepromCrashBegin + SAVE_CRASH_EPC1, epc1);
  109. eepromGet(EepromCrashBegin + SAVE_CRASH_EPC2, epc2);
  110. eepromGet(EepromCrashBegin + SAVE_CRASH_EPC3, epc3);
  111. eepromGet(EepromCrashBegin + SAVE_CRASH_EXCVADDR, excvaddr);
  112. eepromGet(EepromCrashBegin + SAVE_CRASH_DEPC, depc);
  113. snprintf_P(buffer, sizeof(buffer), PSTR("epc1=0x%08x epc2=0x%08x epc3=0x%08x excvaddr=0x%08x depc=0x%08x\n"),
  114. epc1, epc2, epc3, excvaddr, depc);
  115. print.print(buffer);
  116. }
  117. // We need something like
  118. //
  119. //>>>stack>>>
  120. //
  121. //ctx: sys
  122. //sp: 3fffecf0 end: 3fffffb0 offset: 0000
  123. //...addresses...
  124. //...addresses...
  125. //...addresses...
  126. //...addresses...
  127. //<<<stack<<<
  128. //
  129. // Also note that we should always recommend using latest toolchain to decode the .elf,
  130. // older versions (the one used by the 2.3.0 specifically) binutils are broken.
  131. uint32_t stack_start, stack_end;
  132. uint16_t stack_size;
  133. eepromGet(EepromCrashBegin + SAVE_CRASH_STACK_START, stack_start);
  134. eepromGet(EepromCrashBegin + SAVE_CRASH_STACK_END, stack_end);
  135. eepromGet(EepromCrashBegin + SAVE_CRASH_STACK_SIZE, stack_size);
  136. if ((0 == stack_size) || (0xffff == stack_size)) {
  137. return;
  138. }
  139. static constexpr uint16_t StackMin { 0 };
  140. static constexpr uint16_t StackMax { CrashTraceReservedSize };
  141. stack_size = std::clamp(stack_size, StackMin, StackMax);
  142. // offset is technically an unknown, Core's crash handler only gives us `stack_start` as `sp_dump + offset`
  143. // (...maybe we can hack Core / walk the stack / etc... but, that's not really portable between versions)
  144. snprintf_P(buffer, sizeof(buffer),
  145. PSTR("\n>>>stack>>>\n\nctx: todo\nsp: %08x end: %08x offset: 0000\n"),
  146. stack_start, stack_end);
  147. print.print(buffer);
  148. constexpr auto step = sizeof(uint32_t);
  149. int eeprom_addr = EepromCrashBegin + SAVE_CRASH_STACK_TRACE;
  150. uint16_t offset = 0;
  151. uint32_t addr1, addr2, addr3, addr4;
  152. while ((eeprom_addr + (4 * step)) < EepromCrashEnd) {
  153. eepromGet(eeprom_addr, addr1);
  154. eepromGet((eeprom_addr += step), addr2);
  155. eepromGet((eeprom_addr += step), addr3);
  156. eepromGet((eeprom_addr += step), addr4);
  157. snprintf_P(buffer, sizeof(buffer),
  158. PSTR("%08x: %08x %08x %08x %08x \n"),
  159. stack_start + (offset * 4),
  160. addr1, addr2, addr3, addr4
  161. );
  162. print.print(buffer);
  163. eeprom_addr += step;
  164. offset += step;
  165. }
  166. static const char Tail[] PROGMEM = "<<<stack<<<\n";
  167. memcpy_P(buffer, Tail, sizeof(Tail));
  168. print.print(buffer);
  169. }
  170. void forceDump(Print& print) {
  171. dump(print, false);
  172. }
  173. void dump(Print& print) {
  174. dump(print, true);
  175. }
  176. #if TERMINAL_SUPPORT
  177. PROGMEM_STRING(Name, "CRASH");
  178. void command(::terminal::CommandContext&& ctx) {
  179. debug::crash::forceDump(ctx.output);
  180. terminalOK(ctx);
  181. }
  182. static constexpr ::terminal::Command Commands[] PROGMEM {
  183. {Name, command},
  184. };
  185. #endif
  186. } // namespace crash
  187. } // namespace
  188. } // namespace debug
  189. /**
  190. * Save crash information in EEPROM
  191. * This function is called automatically if ESP8266 suffers an exception
  192. * It should be kept quick / consise to be able to execute before hardware wdt may kick in
  193. * This method assumes EEPROM has already been initialized, which is the first thing ESPurna does
  194. */
  195. extern "C" void custom_crash_callback(struct rst_info * rst_info, uint32_t stack_start, uint32_t stack_end ) {
  196. // Small safeguard to protect from calling crash handler very early on boot.
  197. if (!eepromReady()) {
  198. return;
  199. }
  200. // If we crash more than once in a row, don't store (similar) crash log every time
  201. if (systemStabilityCounter() > 1) {
  202. return;
  203. }
  204. // Do not record crash data when doing a normal reboot or when crash trace was disabled
  205. if (pendingDeferredReset()) {
  206. return;
  207. }
  208. if (!debug::crash::enabled()) {
  209. return;
  210. }
  211. // We will use this later as a marker that there was a crash
  212. uint32_t crash_time = millis();
  213. eepromPut(EepromCrashBegin + SAVE_CRASH_CRASH_TIME, crash_time);
  214. // XXX rst_info::reason and ::exccause are uint32_t, but are holding small values
  215. // make sure we are using ::write() instead of ::put(), former tries to deduce the required size based on variable type
  216. eepromWrite(EepromCrashBegin + SAVE_CRASH_RESTART_REASON,
  217. static_cast<uint8_t>(rst_info->reason));
  218. eepromWrite(EepromCrashBegin + SAVE_CRASH_EXCEPTION_CAUSE,
  219. static_cast<uint8_t>(rst_info->exccause));
  220. // write epc1, epc2, epc3, excvaddr and depc to EEPROM as uint32_t
  221. eepromPut(EepromCrashBegin + SAVE_CRASH_EPC1, rst_info->epc1);
  222. eepromPut(EepromCrashBegin + SAVE_CRASH_EPC2, rst_info->epc2);
  223. eepromPut(EepromCrashBegin + SAVE_CRASH_EPC3, rst_info->epc3);
  224. eepromPut(EepromCrashBegin + SAVE_CRASH_EXCVADDR, rst_info->excvaddr);
  225. eepromPut(EepromCrashBegin + SAVE_CRASH_DEPC, rst_info->depc);
  226. // EEPROM size is limited, write as little as possible.
  227. // we definitely want to avoid big stack traces, e.g. like when stack_end == 0x3fffffb0 and we are in SYS context.
  228. // but still should get enough relevant info and it is possible to set needed size at build/runtime
  229. static constexpr uint32_t StackMin { 0 };
  230. static constexpr uint32_t StackMax { CrashTraceReservedSize };
  231. const uint16_t stack_size = std::clamp((stack_end - stack_start), StackMin, StackMax);
  232. eepromPut(EepromCrashBegin + SAVE_CRASH_STACK_START, stack_start);
  233. eepromPut(EepromCrashBegin + SAVE_CRASH_STACK_END, stack_end);
  234. eepromPut(EepromCrashBegin + SAVE_CRASH_STACK_SIZE, stack_size);
  235. // write stack trace to EEPROM and avoid overwriting settings and reserved data
  236. // [EEPROM RESERVED SPACE] >>> ... CRASH DATA ... >>> [SETTINGS]
  237. int eeprom_addr = EepromCrashBegin + SAVE_CRASH_STACK_TRACE;
  238. auto *addr = reinterpret_cast<uint32_t*>(stack_start);
  239. while (EepromCrashEnd > eeprom_addr) {
  240. eepromPut(eeprom_addr, *addr);
  241. eeprom_addr += sizeof(uint32_t);
  242. ++addr;
  243. }
  244. eepromForceCommit();
  245. }
  246. void crashForceDump(Print& print) {
  247. debug::crash::forceDump(print);
  248. }
  249. void crashDump(Print& print) {
  250. debug::crash::dump(print);
  251. }
  252. void crashResetReason(Print& print) {
  253. const auto reason = customResetReason();
  254. bool custom { CustomResetReason::None != reason };
  255. print.printf_P(PSTR("last reset reason: %s\n"), custom
  256. ? customResetReasonToPayload(reason).c_str()
  257. : ESP.getResetReason().c_str());
  258. if (!custom) {
  259. print.printf_P(PSTR("extra info: %s\n"), ESP.getResetInfo().c_str());
  260. }
  261. crashDump(print);
  262. }
  263. size_t crashReservedSize() {
  264. return debug::crash::reserved();
  265. }
  266. void crashClear() {
  267. debug::crash::clear();
  268. }
  269. void crashSetup() {
  270. if (!rtcmemStatus()) {
  271. debug::crash::clear();
  272. }
  273. #if TERMINAL_SUPPORT
  274. espurna::terminal::add(debug::crash::Commands);
  275. #endif
  276. debug::crash::enableFromSettings();
  277. }
  278. #endif // DEBUG_SUPPORT